Continue

Local- and SSH-friendly AI coding assistant; maximum control over model and context.

When to use it

• You want to use local models (Ollama, etc.) or your own API keys without sending code to a vendor's cloud.

• You develop on remote machines via SSH and want the same assistant in your editor.

• You prefer an extension that works inside VS Code or JetBrains with minimal vendor lock-in.

Hands-on

• Prerequisites — VS Code (or supported editor); optional: Ollama or API keys for Claude/OpenAI.

• Install — continue.dev. Configure model (local or API) in Continue settings.

• Runnable example — Build the same small API as in the Cursor/Copilot/Windsurf recipes (see Quick product recipe).

Telegram

Continue does not provide a Telegram gateway. For Telegram + AI, use OpenClaw or the custom bot (playgrounds/telegram-ai-bot/).

Quick product recipe

Same small API with Continue; compare workflow (local/SSH) (~15 min).

1. Open a folder (local or over SSH) in your editor with Continue enabled.
2. Use Continue to create a minimal HTTP API (e.g. one GET endpoint and README). If using local models, notice latency vs cloud.
3. Run and verify. Compare workflow and control vs Cursor, Copilot, or Windsurf.

No separate playground in this repo; the "playground" is any folder and the agent-generated API.